A computationally efficient modelling methodology for field-plates in GaN HEMTs

Jason Hodges, Sayed Ali Albahrani, Sourabh Khandelwal

Research output: Chapter in Book/Report/Conference proceedingConference proceeding contribution

Abstract

Multiple field-plates (FP) are employed in typical GaN HEMTs for improving the device performance. In industry standard models, FP regions are modelled by adopting a full HEMT device model. As a result, a GaN HEMT device with two field plates requires the computation of equations which are the equivalent of three HEMT devices; one for the channel region, and two for the field-plate regions. This causes computational inefficiency. In this paper, we present a simplified field-plate modelling methodology which reduces the computational expense through means of approximations whilst maintaining an excellent degree of accuracy. Using insights from TCAD simulations, we develop simplifications for modelling the FP regions, improving the overall model speed. The developed methodology is validated against the TCAD data showing excellent accuracy and significant model speed improvement over the industry standard model.

Original languageEnglish
Title of host publication2019 IEEE BiCMOS and Compound semiconductor Integrated Circuits and Technology Symposium (BCICTS)
Place of PublicationPiscataway, NJ
PublisherInstitute of Electrical and Electronics Engineers (IEEE)
Number of pages4
ISBN (Electronic)9781728105864
ISBN (Print)9781728105871
DOIs
Publication statusPublished - 2019
Event2nd IEEE BiCMOS and Compound Semiconductor Integrated Circuits and Technology Symposium, BCICTS 2019 - Nashville, United States
Duration: 3 Nov 20196 Nov 2019

Conference

Conference2nd IEEE BiCMOS and Compound Semiconductor Integrated Circuits and Technology Symposium, BCICTS 2019
CountryUnited States
CityNashville
Period3/11/196/11/19

Fingerprint

Dive into the research topics of 'A computationally efficient modelling methodology for field-plates in GaN HEMTs'. Together they form a unique fingerprint.

Cite this